Lightning activity prediction method, device, equipment, storage medium and product

Abstract

The invention discloses a lightning activity prediction method, device and equipment, a storage medium and a product. The method comprises the following steps: acquiring thunder and lightning monitoring data and manual observation data in a current time period; performing data fusion on the lightning monitoring data and the manual observation data to obtain fused lightning data; performing region division on the fused thunder and lightning data to obtain at least one grid region; key parameter extraction is carried out on the fused thunder and lightning data in each grid region to obtain target thunder and lightning data in each grid region; and according to the target thunder and lightning data corresponding to each grid region, performing thunder and lightning activity prediction on the corresponding grid region to obtain a thunder and lightning activity prediction result. According to the technical scheme of the embodiment of the invention, the lightning activity prediction efficiency and prediction accuracy are improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a lightning activity prediction method, device, equipment, storage medium and product. Background Art

[0002] Lightning, a highly destructive weather phenomenon in nature, has always been a major challenge facing human society. Its suddenness and violence may not only cause significant losses to people's lives and property, but also pose a potential threat to important industries such as aviation, electricity, and communications. For this reason, in-depth understanding of lightning activities and accurate monitoring and early warning are particularly important.

[0003] In the context of the rapid development of modern meteorology and information technology, although people already have a series of lightning monitoring methods, there is still a big gap between actual needs and the current state of technology. Although traditional manual weather station observations can provide relatively accurate data to a certain extent, their limitations are becoming increasingly prominent. Limited by the discontinuity of observation time, differences in the professional level of observers, and unstable equipment performance, manual observations are difficult to meet the needs of all-weather, all-angle monitoring.

[0004] Current lightning monitoring methods still have much room for improvement in terms of lightning activity pattern identification and parameter estimation. Lightning activity itself is extremely complex and changeable, which makes it difficult for traditional statistical and analytical methods to deeply explore its internal laws. This limitation not only restricts modern society's comprehensive understanding of lightning activity, but also restricts the accuracy and timeliness of lightning activity prediction to a certain extent. Summary of the invention

[0005] The present invention provides a lightning activity prediction method, device, equipment, storage medium and product to improve the lightning activity prediction efficiency and prediction accuracy.

[0006] According to one aspect of the present invention, a method for predicting lightning activity is provided, the method comprising:

[0007] Obtain lightning monitoring data and manual observation data in the current time period;

[0008] fusing the lightning monitoring data and the manual observation data to obtain fused lightning data;

[0009] Dividing the fused lightning data into regions to obtain at least one grid region;

[0010] Extract key parameters of the fused lightning data within each of the grid areas to obtain target lightning data within each of the grid areas;

[0011] According to the target lightning data corresponding to each of the grid areas, lightning activity prediction is performed on the corresponding grid area to obtain a lightning activity prediction result.

[0012] According to another aspect of the present invention, there is provided a lightning activity prediction device, the device comprising:

[0013] A data acquisition module is used to obtain lightning monitoring data and manual observation data in the current time period;

[0014] A data fusion module, used for fusing the lightning monitoring data and the manual observation data to obtain fused lightning data;

[0015] A region division module, used for dividing the fused lightning data into regions to obtain at least one grid region;

[0016] A parameter extraction module, used to extract key parameters from the fused lightning data in each of the grid areas to obtain target lightning data in each of the grid areas;

[0017] The activity prediction module is used to predict the lightning activity of the corresponding grid area according to the target lightning data corresponding to each grid area, so as to obtain the lightning activity prediction result.

[0018] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0019] at least one processor; and

[0020] a memory communicatively connected to the at least one processor; wherein,

[0021] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the lightning activity prediction method described in any embodiment of the present invention.

[0022] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the lightning activity prediction method described in any embodiment of the present invention when executed.

[0023] According to another aspect of the present invention, a computer program product is provided. The computer program product includes a computer program. When the computer program is executed by a processor, the above-mentioned lightning activity prediction method is implemented.

[0024] The technical solution of the embodiment of the present invention obtains fused lightning data by fusing lightning monitoring data and the artificial observation data; divides the fused lightning data into regions to obtain at least one grid region; extracts key parameters from the fused lightning data in each grid region to obtain target lightning data in each grid region; and predicts lightning activity in the corresponding grid region based on the target lightning data corresponding to each grid region to obtain a lightning activity prediction result. The above technical solution improves the prediction accuracy of lightning activity and the prediction efficiency of lightning activity by fusing lightning data from different data sources and dividing different lightning data into regions, and predicts lightning activity for lightning data in the corresponding region, thereby further effectively improving the effectiveness of lightning monitoring and early warning.

[0025] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0027] Figure 1 is a flow chart of a lightning activity prediction method provided according to Embodiment 1 of the present invention;

[0028] Figure 2 is a flow chart of a lightning activity prediction method provided according to Embodiment 2 of the present invention;

[0029] Figure 3 is a schematic diagram of the structure of a lightning activity prediction device provided according to Embodiment 3 of the present invention;

[0030] Figure 4 It is a structural schematic diagram of an electronic device for implementing the lightning activity prediction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] Embodiment 1

[0034] Figure 1 This is a flowchart of a lightning activity prediction method provided in the first embodiment of the present invention. This embodiment is applicable to the case of real-time prediction of lightning activity. The method can be executed by a lightning activity prediction device. The lightning activity prediction device can be implemented in the form of hardware and / or software. The lightning activity prediction device can be configured in an electronic device. Figure 1 As shown, the method includes:

[0035] S110, obtaining lightning monitoring data and manual observation data in the current time period.

[0036] S120, fusing the lightning monitoring data and the manual observation data to obtain fused lightning data.

[0037] S130, dividing the fused lightning data into regions to obtain at least one grid region.

[0038] S140, extract key parameters from the fused lightning data in each grid area to obtain target lightning data in each grid area.

[0039] S150, predicting lightning activity in the corresponding grid area according to the target lightning data corresponding to each grid area, and obtaining a lightning activity prediction result.

[0040] The lightning monitoring data may be lightning data monitored in real time by a lightning detection device, and may specifically include meteorological data such as the time of occurrence of lightning, the location of lightning, lightning intensity, lightning waveform, lightning frequency, and temperature, humidity, and air pressure in the lightning occurrence area. The artificial observation data may be the original lightning data of an artificial meteorological station, and may specifically include lightning events, lightning types, relative directions or paths, impact ranges, intensity descriptions, and durations.

[0041] Among them, the fused lightning data can be a combination of the continuity of lightning monitoring data and the details of manual observation data, and the fused lightning data has specific continuity, real-time and reliability. The fused lightning data combines the continuity and real-time of automatically monitored lightning monitoring data, and also combines more details and type information contained in manual observation data, thereby making the overall quality, accuracy and comprehensiveness of the fused lightning data higher.

[0042] Exemplarily, the existing data fusion algorithm or data fusion model can be used to fuse the lightning monitoring data and the manual observation data; or, in order to further fully utilize the advantages of both data sources, the characteristics of the two data can be learned and fused based on a multi-head attention mechanism and a recurrent neural network.

[0043] In an optional embodiment, data fusion is performed on lightning monitoring data and manual observation data to obtain fused lightning data, including:

[0044] Step a1: Generate a first eigenvector of lightning monitoring data and a second eigenvector of artificial observation data, and concatenate the first eigenvector and the second eigenvector to obtain a target eigenvector.

[0045] Convert the lightning monitoring data into a feature matrix and get the first feature vector F auto ; and, converting the artificial observation data into a feature matrix to obtain the second eigenvector F manual . Concatenate the first eigenvector and the second eigenvector to obtain the target eigenvector F = [F auto ,F manual ].

[0046] Step a2: input the target feature vector into at least one multi-head attention model to obtain the feature extraction results output by each multi-head attention model.

[0047] The attention weights are calculated using the multi-head attention mechanism (model). For the i-th attention head, the query, key, and value are:

[0048]

[0049] in, and represents the learnable parameter matrix, which is used to calculate the query, key, and value of the i-th attention head respectively.

[0050] Based on the query and key of multiple attention heads, an attention score is calculated to measure the similarity between the query and the key:

[0051]

[0052] Among them, d k Represents the dimension of the key, used to scale the scores to prevent vanishing or exploding gradients.

[0053] Use the softmax function to normalize the attention score and get the attention weight:

[0054] a i =softmax(Score(Q i ,K i ))

[0055] Calculate the output of the i-th attention head according to the self-attention weight:

[0056] Q ai =a i V i

[0057] Step a3: concatenate the feature extraction results to obtain the context feature vector Q a .

[0058] Q a =Concat(Q a1 ,Q a2 ,…,Q aH )W o

[0059] Among them, H represents the number of attention heads, W o represents the learnable parameter matrix.

[0060] Step a4: input the context feature vector, the first feature vector and the second feature vector into a pre-trained data fusion model for fusion processing to obtain fused lightning data.

[0061] The pre-trained data fusion model may be a RNN (Recurrent Neural Network) model.

[0062] Specifically, the context feature vector, the first feature vector, and the second feature vector are input into a pre-trained RNN model, and the RNN model performs data feature fusion processing to obtain fused lightning data.

[0063] It should be noted that since the lightning data collected are distributed in different geographical locations and the time of lightning data collection is different, lightning is greatly affected by geographical location factors. Therefore, lightning activity is affected by factors such as geographical environment in addition to natural weather. Therefore, the fused lightning data can be divided into grids according to regions, and the geographical environment factors in different grid areas are different.

[0064] Exemplarily, the fused lightning data may be divided into regions according to lightning location factors of the fused lightning data to obtain at least one grid region, and each grid region contains the fused lightning data in the region.

[0065] To further improve the accuracy of the division of grid areas, the grid density can be dynamically adjusted according to the intensity and geographical characteristics of lightning activity to optimize data representation. By analyzing the intensity and geographical characteristics of lightning activity, it is possible to determine which areas have more frequent lightning activity. Based on this information, the density of the grid is dynamically adjusted so that finer grids are used in areas with frequent activity and coarser grids are used in areas with less activity. The adaptive grid system can more effectively represent lightning data and provide a basis for subsequent analysis and prediction.

[0066] In an optional embodiment, the fused lightning data is divided into regions to obtain at least one grid region, including:

[0067] Step b1, determining an initial area and dividing the initial area to obtain at least one initial grid area.

[0068] Among them, the initial area can be pre-set by relevant technical personnel according to actual needs.

[0069] Step b2: Determine the lightning activity density index and regional geographical characteristics of each initial grid area based on the fused lightning data and the regional related data corresponding to the initial area.

[0070] Optionally, based on the fused lightning data and the regional related data corresponding to the initial area, the lightning activity density index and regional geographical characteristics of each initial grid area are determined, including:

[0071] Step b21: Determine the lightning activity density index corresponding to each initial grid area according to the data volume of the fused lightning data in the corresponding initial grid area and the initial area of ​​the corresponding initial grid area.

[0072] Specifically, if the initial area of ​​the initial grid area in the ijth direction is recorded as A ij , where i and j represent the initial grid areas in different directions. The number of fused lightning data in the initial grid area in the ijth direction is recorded as N ij , the lightning activity density index I of the initial grid area in the ij direction ij is determined as follows:

[0073]

[0074] The density of the grid is preliminarily determined by determining the number of initial grid networks to be selected in different directions. Therefore, the area A of the initial grid area ij Obtained by dividing the area in different directions.

[0075] Step b22: Determine the regional geographical features corresponding to each initial grid area according to the regional related data corresponding to the corresponding initial grid area.

[0076] Optionally, the region-related data includes an elevation value at at least one elevation measurement point in the corresponding initial grid area, and a height difference and a horizontal distance corresponding to the corresponding initial grid area; accordingly, according to the region-related data corresponding to the corresponding initial grid area, determining the regional geographical features corresponding to each initial grid area respectively includes:

[0077] Determine the elevation standard deviation corresponding to the corresponding initial grid area according to each elevation value; determine the slope factor according to the height difference and horizontal distance corresponding to the corresponding initial grid area; determine the regional geographical features corresponding to each initial grid area according to the elevation standard deviation and slope factor corresponding to the corresponding initial grid area.

[0078] Among them, elevation is one of the important factors to characterize lightning activity. The elevation standard deviation can be used to characterize the elevation factor. For the initial grid area in the ij direction, there is at least one elevation measurement point in the area. Different elevation measurement points correspond to at least one elevation value, which is recorded as Where n represents the number of elevation measurement points.

[0079] For the initial grid area in the ij direction, the mean of each elevation value is determined to obtain the mean elevation value According to the elevation values ​​and the mean elevation, determine the elevation standard deviation corresponding to the initial grid area in the ij direction

[0080]

[0081] The slope factor is determined according to the height difference and horizontal distance corresponding to the corresponding initial grid area. For example, if the height difference of the initial grid area in the ij direction is l ij , the horizontal distance is d ij , then the slope factor of the initial grid area in the ij direction is for:

[0082]

[0083] According to the elevation standard deviation and slope factor corresponding to the corresponding initial grid area, the regional geographical features corresponding to each initial grid area are determined. For the initial grid area in the ij direction, the regional geographical features G of the grid are ijfor:

[0084]

[0085] in, and They represent the weights of the elevation factor and the slope factor respectively. The values ​​depend on the importance of each factor to the lightning activity and can be set in advance by relevant technical personnel.

[0086] Step b3: Determine the time scale change factor according to the current time period.

[0087] Considering seasonal and diurnal changes, the time scale change factor T(t) can be defined. For example, for seasonal changes, a sine function can be used to simulate the time scale change factor T(t):

[0088]

[0089] Among them, t represents the current time period, t0 represents the base time (such as a certain day of a certain year), T0 represents the seasonal change period (such as 365 days a year), and α represents the parameter for adjusting the seasonal change amplitude.

[0090] Step b4: Determine the area of ​​each initial grid area according to the regional geographical characteristics and lightning activity density index of the corresponding initial grid area and based on the time scale variation factor.

[0091] According to the regional geographical characteristics G of the initial grid area in the ij direction ij and Lightning Activity Intensity Index I ij , based on the time scale change factor T(t), adjust the density and shape of the initial grid area. If the area length of the initial grid area in the ij direction is L ij :

[0092]

[0093] Among them, L0 represents the initial side length of the initial grid area, which is obtained by dividing the side lengths of the divided area in different directions according to the grid density. When the intensity of lightning activity increases, the terrain undulation increases, or the time scale change factor increases, the side length of the grid unit will decrease, that is, the grid density will increase. The adjustment of grid density and shape forms a mutually influential relationship with the grid area. This relationship enables more flexible adaptation to different geographical areas and lightning activity conditions in the future, improving the accuracy and efficiency of analysis.

[0094] Step b5: performing region division according to the area of ​​each initial grid region to obtain at least one grid region.

[0095] In order to further improve the efficiency and accuracy of lightning activity prediction, key parameters of the fused lightning data in each grid area can be extracted to obtain the target lightning data in each grid area.

[0096] Exemplarily, a pre-trained deep learning model for extracting key parameters can be used to process and analyze the fused lightning data in each grid area, extract key parameters, and obtain target lightning data. Specifically, a convolutional neural network (CNN) can be used to process and analyze the fused lightning data in each grid area (such as lightning location, occurrence time, intensity, activity mode, activity frequency, density, geographical distribution characteristics, temperature, humidity, air pressure and other meteorological data, mapped to an adaptive grid system). The key parameters that affect lightning activity, such as meteorological data, geographical distribution characteristics, etc., are extracted. The extracted high-precision and high-resolution lightning data can provide important information for subsequent lightning activity predictions.

[0097] The target lightning data in the corresponding grid area is analyzed to obtain the lightning activity prediction results in the corresponding grid area.

[0098] Optionally, according to the target lightning data corresponding to each grid area, a lightning activity prediction is performed on the corresponding grid area to obtain a lightning activity prediction result, including:

[0099] For any grid area, the target lightning data in the grid area is input into the pre-trained lightning activity prediction model to obtain the lightning activity prediction result output by the model. The lightning activity prediction model is obtained by pre-training a pre-built prediction network model based on historical lightning data and historical lightning activity prediction results.

[0100] The prediction network model may be an LSTM (Long Short-Term Memory) model. The historical lightning activity prediction result is the true label value of the corresponding historical lightning data.

[0101] Specifically, historical lightning data and its corresponding historical lightning activity prediction results are input into the LSTM model to obtain the model lightning activity prediction results output by the LSTM model; according to the historical lightning activity prediction results and the model lightning activity prediction results, the loss value is determined based on the preset loss function; the LSTM model is trained based on the loss value until the preset model training end condition is met, and the lightning activity prediction model is obtained. Among them, the model training end condition can be that the loss value reaches a set threshold or the loss value tends to be stable.

[0102] Among them, the lightning activity prediction results may include lightning frequency, lightning intensity and lightning activity hazard level assessment (strong, medium, weak).

[0103] In LSTM, each LSTM unit has a tuple (cell), whose state at time t is recorded as ct, and the tuple can also be regarded as the memory unit of LSTM. The reading and modification of the memory unit in LSTM is achieved by controlling the input gate, forget gate and output gate.

[0104] Specifically, the workflow of the LSTM unit is as follows: At each moment, the LSTM unit receives the current state x through three gates. t The input of the input gate is the current and historical lightning activity data, meteorological data, geographical distribution feature data, etc. After being transformed by nonlinear functions, they are superimposed with the memory cell state processed by the forget gate to form a new memory cell state. Finally, the memory cell state forms the output of the LSTM unit through the operation of nonlinear functions and the dynamic control of the output gate, which can predict the trend of lightning activity and possible risks in the future.

[0105] In the above technical solution, a recurrent neural network (RNN) is used to fuse multivariate lightning data, and then the grid is adaptively adjusted. After that, a pattern recognition method based on a convolutional neural network (CNN) is used to obtain high-precision and high-resolution lightning parameter statistics, thereby establishing a lightning activity prediction based on a long short-term memory network (LSTM).

[0106] The technical solution of the embodiment of the present invention obtains fused lightning data by fusing lightning monitoring data and the artificial observation data; divides the fused lightning data into regions to obtain at least one grid region; extracts key parameters from the fused lightning data in each grid region to obtain target lightning data in each grid region; and predicts lightning activity in the corresponding grid region based on the target lightning data corresponding to each grid region to obtain a lightning activity prediction result. The above technical solution improves the prediction accuracy of lightning activity and the prediction efficiency of lightning activity by fusing lightning data from different data sources and dividing different lightning data into regions, and predicts lightning activity for lightning data in the corresponding region, thereby further effectively improving the effectiveness of lightning monitoring and early warning.

[0107] Embodiment 2

[0108] Figure 2 This is a flow chart of a lightning activity prediction method provided in Embodiment 2 of the present invention. This embodiment provides a preferred example based on the above embodiments.

[0109] like Figure 2 As shown, the method comprises the following specific steps:

[0110] S201. Obtain lightning monitoring data and manual observation data in the current time period.

[0111] S202: Generate a target feature vector based on lightning monitoring data and manual observation data.

[0112] Specifically, a first eigenvector of lightning monitoring data is generated, and a second eigenvector of artificial observation data is generated, and the first eigenvector and the second eigenvector are concatenated to obtain a target eigenvector.

[0113] S203: Generate a context feature vector according to the target feature vector and based on a multi-head attention mechanism.

[0114] Specifically, the target feature vector is input into at least one multi-head attention model to obtain the feature extraction results output by each multi-head attention model; data splicing is performed on each feature extraction result to obtain a context feature vector.

[0115] S204, inputting the context feature vector, the first feature vector and the second feature vector into a pre-trained data fusion model for fusion processing to obtain fused lightning data.

[0116] S205: Determine an initial area and divide the initial area into regions to obtain at least one initial grid region.

[0117] S206. Determine the lightning activity density index and regional geographical characteristics of each initial grid area based on the fused lightning data and the regional related data corresponding to the initial area.

[0118] Specifically, according to the amount of fused lightning data in the corresponding initial grid area and the initial area of ​​the corresponding initial grid area, the lightning activity density index corresponding to each initial grid area is determined; according to the regional related data corresponding to the corresponding initial grid area, the regional geographical features corresponding to each initial grid area are determined.

[0119] Optionally, the region-related data include the elevation value at at least one elevation measurement point in the corresponding initial grid area, and the height difference and horizontal distance corresponding to the corresponding initial grid area; accordingly, based on the region-related data corresponding to the corresponding initial grid area, the regional geographical features corresponding to each initial grid area are determined, including: determining the elevation standard deviation corresponding to the corresponding initial grid area based on each elevation value; determining the slope factor based on the height difference and horizontal distance corresponding to the corresponding initial grid area; determining the regional geographical features corresponding to each initial grid area based on the elevation standard deviation and the slope factor corresponding to the corresponding initial grid area.

[0120] S207. Determine a time scale change factor according to the current time period.

[0121] S208. Determine the area of ​​each initial grid area according to the regional geographical characteristics and lightning activity density index of the corresponding initial grid area and based on the time scale variation factor.

[0122] S209: Perform region division according to the area of ​​each initial grid region to obtain at least one grid region.

[0123] S210 , extract key parameters from the fused lightning data in each grid area to obtain target lightning data in each grid area.

[0124] S211. According to the target lightning data corresponding to each grid area, the lightning activity of the corresponding grid area is predicted to obtain a lightning activity prediction result.

[0125] Specifically, for any grid area, the target lightning data in the grid area is input into a pre-trained lightning activity prediction model to obtain a lightning activity prediction result output by the model; wherein the lightning activity prediction model is obtained by pre-training a pre-constructed prediction network model based on historical lightning data and historical lightning activity prediction results.

[0126] Embodiment 3

[0127] Figure 3 The present invention provides a lightning activity prediction device according to the third embodiment of the present invention. The lightning activity prediction device provided by the embodiment of the present invention can be applied to the situation of real-time prediction of lightning activity. The lightning activity prediction device can be implemented in the form of hardware and / or software, such as Figure 3 As shown, the device specifically includes: a data acquisition module 301, a data fusion module 302, a region division module 303, a parameter extraction module 304 and an activity prediction module 305. Among them,

[0128] The data acquisition module 301 is used to acquire lightning monitoring data and manual observation data in the current time period;

[0129] A data fusion module 302 is used to fuse the lightning monitoring data and the manual observation data to obtain fused lightning data;

[0130] A region division module 303 is used to divide the fused lightning data into regions to obtain at least one grid region;

[0131] A parameter extraction module 304 is used to extract key parameters from the fused lightning data in each of the grid areas to obtain target lightning data in each of the grid areas;

[0132] The activity prediction module 305 is used to predict the lightning activity of the corresponding grid area according to the target lightning data corresponding to each grid area, and obtain the lightning activity prediction result.

[0133] The technical solution of the embodiment of the present invention obtains fused lightning data by fusing lightning monitoring data and the artificial observation data; divides the fused lightning data into regions to obtain at least one grid region; extracts key parameters from the fused lightning data in each grid region to obtain target lightning data in each grid region; and predicts lightning activity in the corresponding grid region based on the target lightning data corresponding to each grid region to obtain a lightning activity prediction result. The above technical solution improves the prediction accuracy of lightning activity and the prediction efficiency of lightning activity by fusing lightning data from different data sources and dividing different lightning data into regions, and predicts lightning activity for lightning data in the corresponding region, thereby further effectively improving the effectiveness of lightning monitoring and early warning.

[0134] Optionally, the data fusion module 302 is specifically used for:

[0135] Generate a first feature vector of the lightning monitoring data, generate a second feature vector of the artificial observation data, and perform data splicing on the first feature vector and the second feature vector to obtain a target feature vector;

[0136] Inputting the target feature vector into at least one multi-head attention model to obtain feature extraction results output by each of the multi-head attention models;

[0137] Performing data splicing on each of the feature extraction results to obtain a context feature vector;

[0138] The context feature vector, the first feature vector and the second feature vector are input into a pre-trained data fusion model for fusion processing to obtain fused lightning data.

[0139] Optionally, the region division module 303 includes:

[0140] A region division unit, used for determining an initial region and performing region division on the initial region to obtain at least one initial grid region;

[0141] A geographical feature determination unit, used to determine the lightning activity density index and regional geographical features of each of the initial grid areas according to the fused lightning data and the area-related data corresponding to the initial area;

[0142] a scale change factor determination unit, configured to determine a time scale change factor according to the current time period;

[0143] A regional area determination unit, used to determine the regional area of ​​each of the initial grid areas according to the regional geographical characteristics and lightning activity density index of the corresponding initial grid area and based on the time scale variation factor;

[0144] The area division unit is used to perform area division according to the area of ​​each of the initial grid areas to obtain at least one grid area.

[0145] Optionally, a geographic feature determination unit comprises:

[0146] A density index determination subunit is used to determine the lightning activity density index corresponding to each of the initial grid areas according to the data volume of the fused lightning data in the corresponding initial grid area and the initial area of ​​the corresponding initial grid area;

[0147] The geographic feature determination subunit is used to determine the regional geographic features corresponding to each of the initial grid areas according to the region-related data corresponding to the corresponding initial grid areas.

[0148] Optionally, the region-related data includes an elevation value at at least one elevation measurement point in the corresponding initial grid area, and a height difference and a horizontal distance corresponding to the corresponding initial grid area; accordingly, the geographic feature determination subunit is specifically used for:

[0149] Determine the elevation standard deviation corresponding to the corresponding initial grid area according to each of the elevation values;

[0150] Determine the slope factor based on the height difference and horizontal distance corresponding to the corresponding initial grid area;

[0151] The regional geographical features corresponding to each of the initial grid areas are determined according to the elevation standard deviation and the slope factor corresponding to the corresponding initial grid areas.

[0152] Optionally, the activity prediction module 305 is specifically used for:

[0153] For any grid area, the target lightning data in the grid area is input into the pre-trained lightning activity prediction model to obtain the lightning activity prediction result output by the model;

[0154] The lightning activity prediction model is obtained by pre-training a pre-built prediction network model based on historical lightning data and historical lightning activity prediction results.

[0155] The lightning activity prediction device provided in the embodiment of the present invention can execute the lightning activity prediction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0156] Embodiment 4

[0157] Figure 4 A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0158] like Figure 4 As shown, the electronic device 40 includes at least one processor 41, and a memory connected to the at least one processor 41, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 to the random access memory (RAM) 43. In the RAM 43, various programs and data required for the operation of the electronic device 40 can also be stored. The processor 41, the ROM 42, and the RAM 43 are connected to each other through a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0159] A number of components in the electronic device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0160] The processor 41 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 41 performs the various methods and processes described above, such as a lightning activity prediction method.

[0161] In some embodiments, the lightning activity prediction method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 40 via the ROM 42 and / or the communication unit 49. When the computer program is loaded into the RAM 43 and executed by the processor 41, one or more steps of the lightning activity prediction method described above may be performed. Alternatively, in other embodiments, the processor 41 may be configured to perform the lightning activity prediction method in any other appropriate manner (e.g., by means of firmware).

[0162] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0163] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0164] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0165] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0166] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0167] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0168] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0169] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting lightning activity, characterized in that: include: Obtain lightning monitoring data and manual observation data in the current time period; fusing the lightning monitoring data and the manual observation data to obtain fused lightning data; Dividing the fused lightning data into regions to obtain at least one grid region; Extract key parameters of the fused lightning data within each of the grid areas to obtain target lightning data within each of the grid areas; According to the target lightning data corresponding to each of the grid areas, lightning activity prediction is performed on the corresponding grid area to obtain a lightning activity prediction result.

2. The method according to claim 1, characterized in that The step of fusing the lightning monitoring data and the manual observation data to obtain fused lightning data includes: Generate a first feature vector of the lightning monitoring data, generate a second feature vector of the artificial observation data, and perform data splicing on the first feature vector and the second feature vector to obtain a target feature vector; Inputting the target feature vector into at least one multi-head attention model to obtain feature extraction results output by each of the multi-head attention models; Performing data splicing on each of the feature extraction results to obtain a context feature vector; The context feature vector, the first feature vector and the second feature vector are input into a pre-trained data fusion model for fusion processing to obtain fused lightning data.

3. The method according to claim 1, characterized in that The step of dividing the fused lightning data into regions to obtain at least one grid region includes: Determine an initial region and divide the initial region into regions to obtain at least one initial grid region; Determine the lightning activity density index and regional geographical characteristics of each of the initial grid areas according to the fused lightning data and the area-related data corresponding to the initial area; Determining a time scale change factor according to the current time period; Determine the area of ​​each of the initial grid areas according to the regional geographical characteristics and lightning activity density index of the corresponding initial grid area and based on the time scale variation factor; Region division is performed according to the region area of ​​each of the initial grid regions to obtain at least one grid region.

4. The method according to claim 3, characterized in that Determining the lightning activity density index and regional geographical features of each of the initial grid areas according to the fused lightning data and the area-related data corresponding to the initial area includes: Determine the lightning activity density index corresponding to each of the initial grid areas according to the data volume of the fused lightning data in the corresponding initial grid area and the initial area of ​​the corresponding initial grid area; According to the region-related data corresponding to the corresponding initial grid areas, the regional geographical features respectively corresponding to the initial grid areas are determined.

5. The method according to claim 4, characterized in that The area-related data includes the elevation value of at least one elevation measurement point in the corresponding initial grid area, and the height difference and horizontal distance corresponding to the corresponding initial grid area; Accordingly, the determining of the regional geographical features corresponding to each of the initial grid areas according to the regional related data corresponding to the corresponding initial grid areas includes: Determine the elevation standard deviation corresponding to the corresponding initial grid area according to each of the elevation values; Determine the slope factor based on the height difference and horizontal distance corresponding to the corresponding initial grid area; The regional geographical features corresponding to each of the initial grid areas are determined according to the elevation standard deviation and the slope factor corresponding to the corresponding initial grid areas.

6. The method according to claim 1, characterized in that The step of predicting lightning activity in the corresponding grid area according to the target lightning data corresponding to each grid area to obtain a lightning activity prediction result includes: For any grid area, the target lightning data in the grid area is input into the pre-trained lightning activity prediction model to obtain the lightning activity prediction result output by the model; The lightning activity prediction model is obtained by pre-training a pre-built prediction network model based on historical lightning data and historical lightning activity prediction results.

7. A lightning activity prediction device, characterized in that: include: A data acquisition module is used to obtain lightning monitoring data and manual observation data in the current time period; A data fusion module, used for fusing the lightning monitoring data and the manual observation data to obtain fused lightning data; A region division module, used for dividing the fused lightning data into regions to obtain at least one grid region; A parameter extraction module, used to extract key parameters from the fused lightning data in each of the grid areas to obtain target lightning data in each of the grid areas; The activity prediction module is used to predict the lightning activity of the corresponding grid area according to the target lightning data corresponding to each grid area, so as to obtain the lightning activity prediction result.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the lightning activity prediction method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the lightning activity prediction method according to any one of claims 1 to 6 when executed.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the computer program implements the lightning activity prediction method according to any one of claims 1 to 6.

Images